Warning of AI’s danger, pioneer Geoffrey Hinton quits Google to speak freely

In March, as part of the company's plan to lay off 10,000 employees, it shut down its Ethics and Society team. Today, in a blog post, Natasha Crampton, Microsoft's Chief Responsible AI Officer, wrote about why that decision was made as part of an overall look into the company's responsible AI plans.

Crampton stated that the Ethics and Society team "was important to enabling us to get where we are today." However, she said that Microsoft decided to make some organizational changes for the company's responsible AI division:

Last year, we made two key changes to our responsible AI ecosystem: first, we made critical new investments in the team responsible for our Azure OpenAI Service, which includes cutting-edge technology like GPT-4; and second, we infused some of our user research and design teams with specialist expertise by moving former Ethics & Society team members into those teams. Following those changes, we made the hard decision to wind down the remainder of the Ethics & Society team, which affected seven people. No decision affecting our colleagues is easy, but it was one guided by our experience of the most effective organizational structures to ensure our responsible AI practices are adopted across the company.

Crampton says that Microsoft now has 350 people working in some capacity on responsible AI. Of that number, 129 employees are working full-time on just that topic and the rest have responsible AI as a major part of their duties. She added that Microsoft has plans to hire new employees and move some current ones to work full-time on responsible AI duties, and more will be revealed on those efforts "in the coming months."

The blog post also talks about Microsoft's Responsible AI Council, which meets regularly to discuss these topics. The council is chaired by the company's president Brad Smith and its CTO Kevin Scott. It also talks about how Microsoft is working to put in responsible AI ethics across all of the company's divisions. Crampton stated:

We have senior leaders tasked with spearheading responsible AI within each core business group and we continually train and grow a large network of responsible AI “champions” with a range of skills and roles for more regular, direct engagement.

While all of these plans sound like Microsoft is taking the concerns of many about the rise of AI seriously, we will have to see if the company's actions will be the same as its words in the months and years ahead.

Microsoft goes into more detail on how it plans to regulate responsible AI products

OpenAI gives in to Italy’s data privacy demands, ending ChatGPT ban

Going back to the popularity aspect of large displays, Technavio reports that just five years ago, 40-inch TVs were the standard size. This is rising, though, because manufacturers now offer TVs in excess of 80 inches. Another reason people are buying 4K TVs is due to their smart features. People tend to buy new TVs every decade or so, this means some people may not have built-in smart features yet but want them.

The main key challenge to the growth of 4K TVs is expected to be the lack of 4K content. While these TVs can play 1080p and 720p content, consumers may be able to spend less money on a less sophisticated TV if they don’t think 4K is worth it. As time rolls on, this should become less of an issue as 4K-capable recording equipment comes down in price and more content is produced.

While over the whole period, a CAGR of 19.19% is certainly possible, in the short term figures could be lower. Many sectors, including semiconductors and smartphones, have seen falling revenues due to the poor economic situation.

Let us know in the comments if you’re planning to upgrade to a 4K TV soon. What do you think are the pros and cons right now?

The 4K TV market will see a 19.19% compound annual growth rate until 2027

Source

Pressed by Republican senators to sanction Chinese cloud computing companies, US Commerce Secretary Gina Raimondo made it clear that she and the Biden administration would not be painted as weak on China:

“I’ve put over 200 Chinese companies on the entity list in my tenure,” she told the Senate Appropriations Committee, “and we are actively, constantly investigating additional threats and if and as we think companies need to go onto the list, I will not hesitate.”

On April 25, Senator Bill Hagerty (R-TN), ranking member of the Senate Banking Subcommittee on National Security and International Trade and Finance, and eight other Republican senators sent a letter to Raimondo, Treasury Secretary Janet Yellen and Secretary of State Antony Blinken asking them to take decisive action against Alibaba Cloud, Huawei Cloud and other Chinese cloud services providers including Baidu and Tencent:

“Open-source information shows that Huawei Cloud and PRC-based cloud computing services not only pose a threat to US national security and economic security, but also are increasingly engaging with foreign entities—in some cases sanctioned foreign entities—that are directly challenging the national security and economic security interests of the United States and our allies and partners. We urge you to use all available tools to engage in decisive action against these firms, through sanctions, export restrictions, and investment bans, and to further investigate PRC cloud computing service companies,” they wrote.

An exchange posted on Senator Hagerty’s website reads:

“The examples that we provide in the letter show that China’s military civil fusion strategy is alive and well. Given China’s laws that compel every Chinese citizen and Chinese company to assist in national security or intelligence work, it should be common sense to deny US exports of these countries,” Hagerty said. “Yet companies like Alibaba Cloud that are not on the Entity List and still have access to US technology and exports and even have operations here in the United States.”

“Given the far reach of China’s national security related laws and their military civil fusion strategy, do you agree that the PRC cloud service providers that operate in the United States pose a threat to our national and our economic security?” Hagerty asked.

“I’m in broad agreement with you,” said Secretary Raimondo.

Secretary of Commerce Gina Raimondo. Photo: Asia Times files

The letter goes on to note specific but previously known examples of the companies’ work with China’s military, security and intelligence services, including domestic monitoring and satellite imagery. However, like military civil fusion company Boeing in the US, they primarily serve commercial markets.

For reference, “military civil fusion” is a common practice, not a nefarious Chinese plot. In February, Boeing announced that it had been “selected by the US Air Force as the prime contractor for the nation’s intercontinental ballistic missile (ICBM) guidance subsystems support. The contract is worth up to $1.6 billion over 16 years.”

Alibaba Cloud is the fourth largest cloud service provider in the world and the largest in China. It provides database, storage, data analytics, networking, application, security and other services to the finance, e-commerce, logistics and other sectors worldwide.

Employment website indeed.com currently lists 30 positions available at Alibaba in Sunnyvale, California, including software engineer, photonics expert, research scientist and business manager, with annual compensation ranging from $110,000 to $240,000.

Crunchbase estimates that tech companies in the US have cut more than 135,000 jobs so far this year. Senator Hagerty and his colleagues would presumably like to eliminate jobs at Alibaba as well – and the technological expertise they represent.

The level of expertise is high. For example, candidates for the position of photonics expert must have a Ph.D. degree in electrical engineering, applied physics or a related field, and experience in the design, development and fabrication of silicon photonics chips for optical interconnects.

Huawei Cloud is only about half the size of Alibaba Cloud, but it has been advancing rapidly. Since being added to the Entity List in August 2020, Huawei Cloud, as reported on its website, has received several certifications and awards, including:

August 2020 – The first cloud service provider in Asia-Pacific to earn the PCI 3DS certification for global payment account data security.

September 2020 – The first cloud service provider to earn the ISO 27799 certification for information security management.

November 2021 – Awarded both the ISO/IEC 27034 application security standard certification and the CSA STAR 2021 Gold Certification by the British Standards Institution.

November 2020 – Listed as a leader in predictive analytics and machine learning by Forrester Research.

December 2020 – Placed by Gartner in itsMagic Quadrant for Cloud Database Management Systems for its GaussDB database products, which help enterprises migrate databases to the cloud.

March 2021 – Named by Canalys as one of the top four and as the fastest growing cloud service provider in Latin America.

May 2021 – Ranked No. 1 in China’s financial cloud infrastructure market by IDC.

December 2021 – Ranked No.1 in China’s public cloud service market for machine learning by IDC.

In March 2022, the Huawei Cloud Summit Middle East and Africa was held in Dubai with more than 1,000 government officials, customer representatives and industry experts attending.

The conference considered how cloud computing can be used to facilitate public services, finance, carriers, media, e-commerce and gaming.

In September 2022, the first Huawei Cloud Indonesia Summit was held in Jakarta with local industry leaders and experts, business partners and the media attending to discuss the digital economy.

In November, Indonesia became a new region for Huawei Cloud with more than 60 services to be launched in areas including e-commerce, short video platforms, online gaming insurance and finance.

In November 2020, Huawei Cloud and BCB Blockchain of Singapore established a partnership to develop smart city solutions.

In September 2022, the Huawei Cloud Accelerator, a program aimed at promoting the development of start-ups with the help of venture capitalists, was announced at the Huawei Cloud Global Startup Founders Summit held the Huawei Cloud Accelerator, a program aiming at empowering startups in Shenzhen.

In China, the company reports that more than 800 government clouds are now built on Huawei Cloud, the company’s “One City, One Cloud” concept has been implemented in more than 150 cities, and more than 300 financial institutions including six major banks, 90% of the top 50 e-commerce companies and 90% of the top 30 auto companies use Huawei Cloud.

Other user industries include e-commerce, finance, energy, manufacturing, pharmaceuticals and gaming. Outside China, Huawei Cloud has a presence in the Asia-Pacific, the Middle East, Africa, Latin America and Europe. Aside from not being able to work in the US, being on the Entity List does not seem to be a major problem.

Will US politicians be able to throw a spanner into the works of other Chinese cloud service providers? Perhaps. They could force US companies including Salesforce, IBM, VMware and Fortinet to sever their ties with Alibaba Cloud, just as they forced Oracle to sever its ties with Huawei.

That prompted Huawei to develop its own ERP software, which it announced in April. Something similar would probably happen if Alibaba Cloud were put on the Entity List managed by the Commerce Department’s Bureau of Industry and Security (BIS).

Being put on the list can be a serious problem, but it also forces the targeted company to try harder. In the Global South and other parts of the world where US sanctions are not recognized, it might be called the BIS Certificate of Quality.

Source

Translating someone’s brain activity into written words may sound like a science fiction dream, but a new artificial intelligence (AI) model developed at the University of Texas at Austin has been able to achieve just that. Using only noninvasive scanning methods, the model can be trained to decode complex language from someone’s thoughts for extended periods of time.

“For a noninvasive method, this is a real leap forward compared to what’s been done before, which is typically single words or short sentences,” said study co-lead Alex Huth, an assistant professor of neuroscience and computer science, in a statement.

Other similar systems are in development elsewhere, but what sets this one apart is that participants don’t need to undergo surgery to have implants fitted, nor are they restricted to a list of words they can use. 

Using technology like that seen in Open AI’s ChatGPT and Google’s Bard chatbots, the model – called a semantic decoder – is trained on hours of data obtained from an individual as they listen to podcasts whilst having their brain scanned via functional magnetic resonance imaging (fMRI). Later, with the participant’s consent, they can have their thoughts decoded while listening to a new story or imagining telling a story, and the model will generate a stream of text.

The results looked something like this:

The stories the participants were listening to are shown on the left; on the right is what the model was able to decode from their brain activity. Image credit: University of Texas at Austin

The decoder can’t synthesize the person’s thoughts word for word, but it can often capture the gist of what they’re thinking. After extensive training, it’s able to produce text that is a good, and occasionally exact, representation of the person’s thoughts about half of the time. 

The study wasn’t just limited to hearing or thinking about stories. This video shows what the model was able to decode from someone’s brain activity while they were watching a movie clip with the sound turned off:

It may not be perfect, but the fact that the whole process is noninvasive is a big plus. In the future, it’s hoped that further development of technology like this could help patients who are no longer able to physically communicate via speech, such as some stroke survivors.

But if looking at this kind of tech gives you an uneasy feeling, you’re not alone. For many people, a device that can read your thoughts is more the stuff of dystopian nightmares than sci-fi fantasy. 

Addressing these inevitable fears, study co-lead and doctoral student Jerry Tang said, “We take very seriously the concerns that it could be used for bad purposes and have worked to avoid that. We want to make sure people only use these types of technologies when they want to and that it helps them.”

For starters, there’s the practical consideration that this system has to be trained for hours before it can begin to work. “A person needs to spend up to 15 hours lying in an MRI scanner, being perfectly still, and paying good attention to stories that they’re listening to before this really works well on them,” Huth explained.

Beyond that, there’s a failsafe: even someone who had participated in training the model could prevent it from decoding their inner speech by thinking of something unrelated, such as animals.

Still, as the researchers continue to work to progress this technology, privacy and safety are at the forefront. “I think right now, while the technology is in such an early state, it’s important to be proactive by enacting policies that protect people and their privacy,” said Tang. “Regulating what these devices can be used for is also very important.”

The study is published in Nature Neuroscience.

Source

A new bill proposed by US lawmakers would prevent artificial intelligence (AI) from being able to singlehandedly launch nuclear weapons without human input, codifying existing Pentagon rules.

While current rules forbid the autonomous launch of nuclear weapons, there aren’t any actual laws that prevent this from happening. With the astronomical rise in AI models in recent years, officials have become concerned that they could sneak their way into the very top-level decision-making of the US military. 

In anticipation of this possibility, Senator Edward Markey (D-Mass.) and Representatives Ted W. Lieu (CA-36), Don Beyer (VA-08), and Ken Buck (CO-04) have introduced a bipartisan bill called Block Nuclear Launch by Autonomous Artificial Intelligence Act that will “safeguard the nuclear command and control process from any future change in policy that allows artificial intelligence (AI) to make nuclear launch decisions”.

It will ensure humans are “in the loop” following an order by the President to launch a nuclear weapon, for either defense or offense. 

“AI technology is developing at an extremely rapid pace,” said Representative Ted Lieu in a statement.

“While we all try to grapple with the pace at which AI is accelerating, the future of AI and its role in society remains unclear. It is our job as Members of Congress to have responsible foresight when it comes to protecting future generations from potentially devastating consequences. That’s why I’m pleased to introduce the bipartisan, bicameral Block Nuclear Launch by Autonomous AI Act, which will ensure that no matter what happens in the future, a human being has control over the employment of a nuclear weapon – not a robot. AI can never be a substitute for human judgment when it comes to launching nuclear weapons."

The bill is following-through on a recommendation from a 2021 National Security Commission on Artificial Intelligence report that suggested such a law, in the hopes that the US would spearhead the idea for other nuclear powers to follow.  

AI models have no concept of empathy and would not truly understand the impact of a nuclear weapon, so allowing them uncontrolled access to the launch systems could lead to a disaster that could otherwise be averted. For example, Soviet submariner Vasili Arkhipov single-handedly prevented nuclear war when their Captain mistakenly thought that war had broken out between the US and the Soviet Union – had AI been at the helm, the world could look very different today. 

Source

The NVIDIA GeForce RTX 4060 Ti is set to be unveiled and subsequently hit the shelves by the "end of May," as per a snippet from a leaked non-disclosure agreement (NDA) obtained by VideoCardz. The exact date may be finalized closer to the launch. The leaked document reveals NVIDIA is already sending RTX 4060 Ti units to influencers and reviewers in media and will start shipping stocks to channel partners by May 5.

Meanwhile, the launch timeline for the RTX 4060 has yet to be confirmed. It may be expected to launch before NVIDIA shifts its focus to the rumored RTX 4080 Ti.

Rumored NVIDIA GeForce 4060 Ti specifications and pricing

The rumored specifications of the NVIDIA GeForce 4060 Ti first surfaced online in December last year, suggesting a minor upgrade over the predecessor, RTX 3060 Ti. It reportedly uses NVIDIA's Ada Lovelace architecture and utilizes 4352 CUDA cores, with an 8GB GDDR6 video memory.

The card has a rumored speed of 18Gbps and a 288GB/s bandwidth over a 128-bit wide bus. The GPU is also said to feature a 32MG L2 cache, which is the same as the laptop version of the RTX 4060.

The RTX 4060 Ti is rumored to be rated at 160W in terms of TDP, as per leaker @kopite7kimi, leading to significantly lower than the RTX 3060 Ti. Notably, NVIDIA is still sticking to the same PG190 board for the RTX 4060 Ti as the previous generation, which may, paired with the lower TGP, suggest minor performance improvements.

However, what may come as a delight to buyers is the rumored price. According to YouTuber RedGamingTech, the GeForce RTX 4060 Ti may get a price of just $399. As per the creator, the MSRP has been slashed after witnessing poor demand for the RTX 4070.

Meanwhile, the GeForce RTX 4060 gets a VRAM downgrade from the RTX 3060 and is expected to offer a much lower TGP of around 115W, fewer cores, and a smaller L2 cache as compared to the RTX 4060 Ti. While it can be expected to retail at an even lower price, enthusiasts will have to wait longer for an official word from NVIDIA on that.

NVIDIA GeForce RTX 4060 Ti launch timeline and pricing tipped in separate leaks

Microsoft has announced a redesigned dashboard for Xbox Insiders in the Alpha Skip-Ahead and Alpha Ring. The latest version addresses feedback Microsoft gathered during its months-long experiments with the Home screen, such as space issues with a new dashboard variant and an overly crammed UI layout.

Here is what the revised variant will offer this week to those testing early Xbox updates:

• Simplified navigation with better access to the Microsoft Store, Xbox Game Pass, search, and settings. All those options are now available at the top of the Home screen.
• Decluttered dashboard layout with more emphasis and space for your backgrounds. Some tiles are now smaller, while others Microsoft moved to the bottom of the screen.
• A new responsive art feature will change your dashboard background to show off beautiful art associated with each game.
• The "My games & apps" tile now can let you know if an app or game needs your attention.

Microsoft has added that the announced changes do not represent the final experience, so Xbox Insiders can expect more tweaks and refinements before the public rollout. You can learn more about the latest dashboard changes in the Xbox Insider program in a blog post on the official website.

Microsoft announces a new and simplified Home experience for Xbox Insiders

Thirty years ago, Tim Berners-Lee and CERN gave the world a gift. After some internal discussions, CERN stamped (quite literally) its approval on a document, relinquishing all intellectual property rights to the World Wide Web and entering it into the public domain. It wasn’t a decision that was altogether altruistic—there were certainly logistical and operational reasons for the choice—but it was a decision grounded in optimism and ambition. It radically changed the trajectory of the web, enabling an openness and accessibility that most technologies don’t come close to reaching.

And yet, the most shocking part of the decision was that it wasn’t shocking at all.

Berners-Lee and his team at CERN, together with web pioneers from around the world, had a lot of decisions to make early on. Because the web is globally distributed and publicly accessible, those decisions could rarely be reversed. A lot of time was given to the formation of syntax, the language of protocols and specifications, the design and details of the browser, and dozens of other things. The early web community was making decisions each and every day, one more deliberate and precise than the next.

In 1999, Tim Berners-Lee published his own history building the web in a book, titled Weaving the Web in which he describes many of these decisions at great length. But CERN’s decision to enter the web into the public domain gets only a few paragraphs. There were some discussions, and there was a chance that maybe it didn’t happen, but all in all it happened relatively quickly. It was a product of timing and position.

For a while, in the early 90’s, the web was in competition with Gopher, an alternative Internet-backed protocol developed at the University of Minnesota, and named for the school’s mascot. Gopher shared some philosophical preferences with the web, but it was, in practice, something entirely different. Gopher was hierarchical by design, a stark contrast to the sprawling, nonhierarchic hyperlink of the web. Gopher also put resources behind search and design of sites years before technologies like Google and browsers like Netscape would do the same for the web.

As a result, it was pretty popular. Maybe more so than the web. But there’s a reason we don’t hear much about Gopher these days.

In February of 1993, the University of Minnesota made an announcement. In specific commercial usage of the protocol, they would be charging licensing fees. Not large fees, and not in all cases. But, in some small way, they would be restricting access.

Overnight, sentiment shifted. Internet users took to BBS boards and mailing lists to express outrage about Gopher’s decision. IBM declared that they wouldn’t support internally any protocol with restrictive licensing. The world began searching for alternatives.

Back at CERN, Berners-Lee had just the thing. He was already in early conversations with CERN about an open-source license for the platform. But after the Gopher announcement, he changed that request. He didn’t want any restrictions at all, no rights or attachments at all. The web needed to be free. Truly free.

CERN had some advantages that the University of Minnesota didn’t. Financially there wasn’t a ton to be gained from maintaining such a massive technology so far outside of their core competency (they’re a particle physics lab after all). And in terms of scale, they’re operating an order of magnitude or two above any research wing of a university. So a technology like the web, as it big as it would become, wasn’t as important to their organization or its mission. Besides, they saw what happened at Gopher and didn’t want to deal with that kind of fallout.

But also, and this is important to mention, they believed in the web and in Berners-Lee. The folks making these decisions understood its potential and wanted the web to flourish. This wasn’t a decision driven by profit. It was a generous and enthusiastic vote of confidence in the global ambitions of the web.

And so, on April 30 1995, with a half-page document and a few signatures, CERN made it official. Nobody owned the web, and the web wasn’t licensed. It was simply a part of the world, for anybody to use, distribute, or modify. It looked like this:

Recalling that decision ten years later, Tim Berners-Lee recorded a short message about the continued openness of the web:

If we did not have that document from CERN, we would not have the web. But the web’s not done.

And it’s still not done.

Source

Geoffrey Hinton was an artificial intelligence pioneer. In 2012, Dr. Hinton and two of his graduate students at the University of Toronto created technology that became the intellectual foundation for the A.I. systems that the tech industry’s biggest companies believe is a key to their future.

On Monday, however, he officially joined a growing chorus of critics who say those companies are racing toward danger with their aggressive campaign to create products based on generative artificial intelligence, the technology that powers popular chatbots like ChatGPT.

Dr. Hinton said he has quit his job at Google, where he has worked for more than decade and became one of the most respected voices in the field, so he can freely speak out about the risks of A.I. A part of him, he said, now regrets his life’s work.

“I console myself with the normal excuse: If I hadn’t done it, somebody else would have,” Dr. Hinton said during a lengthy interview last week in the dining room of his home in Toronto, a short walk from where he and his students made their breakthrough.

Dr. Hinton’s journey from A.I. groundbreaker to doomsayer marks a remarkable moment for the technology industry at perhaps its most important inflection point in decades. Industry leaders believe the new A.I. systems could be as important as the introduction of the web browser in the early 1990s and could lead to breakthroughs in areas ranging from drug research to education.

But gnawing at many industry insiders is a fear that they are releasing something dangerous into the wild. Generative A.I. can already be a tool for misinformation. Soon, it could be a risk to jobs. Somewhere down the line, tech’s biggest worriers say, it could be a risk to humanity.

“It is hard to see how you can prevent the bad actors from using it for bad things,” Dr. Hinton said.

After the San Francisco start-up OpenAI released a new version of ChatGPT in March, more than 1,000 technology leaders and researchers signed an open letter calling for a six-month moratorium on the development of new systems because A.I technologies pose “profound risks to society and humanity.”

Several days later, 19 current and former leaders of the Association for the Advancement of Artificial Intelligence, a 40-year-old academic society, released their own letter warning of the risks of A.I. That group included Eric Horvitz, chief scientific officer at Microsoft, which has deployed OpenAI’s technology across a wide range of products, including its Bing search engine.

Dr. Hinton, often called “the Godfather of A.I.,” did not sign either of those letters and said he did not want to publicly criticize Google or other companies until he had quit his job. He notified the company last month that he was resigning, and on Thursday, he talked by phone with Sundar Pichai, the chief executive of Google’s parent company, Alphabet. He declined to publicly discuss the details of his conversation with Mr. Pichai.

Google’s chief scientist, Jeff Dean, said in a statement: “We remain committed to a responsible approach to A.I. We’re continually learning to understand emerging risks while also innovating boldly.”

Dr. Hinton, a 75-year-old British expatriate, is a lifelong academic whose career was driven by his personal convictions about the development and use of A.I. In 1972, as a graduate student at the University of Edinburgh, Dr. Hinton embraced an idea called a neural network. A neural network is a mathematical system that learns skills by analyzing data. At the time, few researchers believed in the idea. But it became his life’s work.

In the 1980s, Dr. Hinton was a professor of computer science at Carnegie Mellon University, but left the university for Canada because he said he was reluctant to take Pentagon funding. At the time, most A.I. research in the United States was funded by the Defense Department. Dr. Hinton is deeply opposed to the use of artificial intelligence on the battlefield — what he calls “robot soldiers.”

In 2012, Dr. Hinton and two of his students in Toronto, Ilya Sutskever and Alex Krishevsky, built a neural network that could analyze thousands of photos and teach itself to identify common objects, such as flowers, dogs and cars.

Google spent $44 million to acquire a company started by Dr. Hinton and his two students. And their system led to the creation of increasingly powerful technologies, including new chatbots like ChatGPT and Google Bard. Mr. Sutskever went on to become chief scientist at OpenAI. In 2018, Dr. Hinton and two other longtime collaborators received the Turing Award, often called “the Nobel Prize of computing,” for their work on neural networks.

Ilya Sutskever, OpenAi’s chief scientist, worked with Dr. Hinton on his research in Toronto.Credit...Jim Wilson/The New York Times

Around the same time, Google, OpenAI and other companies began building neural networks that learned from huge amounts of digital text. Dr. Hinton thought it was a powerful way for machines to understand and generate language, but it was inferior to the way humans handled language.

Then, last year, as Google and OpenAI built systems using much larger amounts of data, his view changed. He still believed the systems were inferior to the human brain in some ways but he thought they were eclipsing human intelligence in others. “Maybe what is going on in these systems,” he said, “is actually a lot better than what is going on in the brain.”

As companies improve their A.I. systems, he believes, they become increasingly dangerous. “Look at how it was five years ago and how it is now,” he said of A.I. technology. “Take the difference and propagate it forwards. That’s scary.”

Until last year, he said, Google acted as a “proper steward” for the technology, careful not to release something that might cause harm. But now that Microsoft has augmented its Bing search engine with a chatbot — challenging Google’s core business — Google is racing to deploy the same kind of technology. The tech giants are locked in a competition that might be impossible to stop, Dr. Hinton said.

His immediate concern is that the internet will be flooded with false photos, videos and text, and the average person will “not be able to know what is true anymore.”

He is also worried that A.I. technologies will in time upend the job market. Today, chatbots like ChatGPT tend to complement human workers, but they could replace paralegals, personal assistants, translators and others who handle rote tasks. “It takes away the drudge work,” he said. “It might take away more than that.”

Down the road, he is worried that future versions of the technology pose a threat to humanity because they often learn unexpected behavior from the vast amounts of data they analyze. This becomes an issue, he said, as individuals and companies allow A.I. systems not only to generate their own computer code but actually run that code on their own. And he fears a day when truly autonomous weapons — those killer robots — become reality.

“The idea that this stuff could actually get smarter than people — a few people believed that,” he said. “But most people thought it was way off. And I thought it was way off. I thought it was 30 to 50 years or even longer away. Obviously, I no longer think that.”

Many other experts, including many of his students and colleagues, say this threat is hypothetical. But Dr. Hinton believes that the race between Google and Microsoft and others will escalate into a global race that will not stop without some sort of global regulation.

But that may be impossible, he said. Unlike with nuclear weapons, he said, there is no way of knowing whether companies or countries are working on the technology in secret. The best hope is for the world’s leading scientists to collaborate on ways of controlling the technology. “I don’t think they should scale this up more until they have understood whether they can control it,” he said.

Dr. Hinton said that when people used to ask him how he could work on technology that was potentially dangerous, he would paraphrase Robert Oppenheimer, who led the U.S. effort to build the atomic bomb: “When you see something that is technically sweet, you go ahead and do it.”

He does not say that anymore.

Source

According to ChatGPT, it was July 10, 1956, in an article titled “Machines Will Be Capable of Learning, Solving Problems, Scientists Predict” about a seminal conference at Dartmouth College. The chatbot added:

CHATGPT

This conference is now considered to be the birthplace of artificial intelligence as a field of study, and the article mentions the term “artificial intelligence” several times in the context of the discussions and presentations that took place at the conference.

The 1956 conference was real. The article was not. ChatGPT simply made it up. ChatGPT doesn’t just get things wrong at times, it can fabricate information. Names and dates. Medical explanations. The plots of books. Internet addresses. Even historical events that never happened.

When ChatGPT was recently asked how James Joyce and Vladimir Lenin first met — there is no evidence they ever did — this is how it responded:

CHATGPT

James Joyce and Vladimir Lenin met in Zurich, Switzerland in 1916. Both men were living in exile in Zurich during World War I. Joyce was a writer and Lenin was a revolutionary. They met at the Cafe Odéon, a popular gathering place for artists and intellectuals in Zurich.

The response above has been truncated for length.

Fabrications like these are common. Figuring out why chatbots make things up and how to solve the problem has become one of the most pressing issues facing researchers as the tech industry races toward the development of new A.I. systems.

Chatbots like ChatGPT are used by hundreds of millions of people for an increasingly wide array of tasks, including email services, online tutors and search engines. And they could change the way people interact with information. But there is no way of ensuring that these systems produce information that is accurate.

The technology, called generative A.I., relies on a complex algorithm that analyzes the way humans put words together on the internet. It does not decide what is true and what is not. That uncertainty has raised concerns about the reliability of this new kind of artificial intelligence and calls into question how useful it can be until the issue is solved or controlled.

The tech industry often refers to the inaccuracies as “hallucinations.” But to some researchers, “hallucinations” is too much of a euphemism. Even researchers within tech companies worry that people will rely too heavily on these systems for medical and legal advice and other information they use to make daily decisions.

“If you don’t know an answer to a question already, I would not give the question to one of these systems,” said Subbarao Kambhampati, a professor and researcher of artificial intelligence at Arizona State University.

ChatGPT wasn’t alone in erring on the first reference to A.I. in The Times. Google’s Bard and Microsoft’s Bing chatbots both repeatedly provided inaccurate answers to the same question. Though false, the answers seemed plausible as they blurred and conflated people, events and ideas.

Google’s Bard said:

GOOGLE’S BARD

The earliest mention of artificial intelligence (AI) in The New York Times was in 1956, when the term was coined by John McCarthy 1 at a workshop at Dartmouth College. The workshop brought together some of the brightest minds in computer science to discuss the possibility of creating machines that could think like humans.

The Times reported on the workshop in a front-page article 2, calling it “a new field of science” 3 that was “aimed at creating machines that can think like humans.” The article quoted McCarthy as saying 4 that the goal of AI was to “build machines with minds, capable of learning and adapting to new situations.”

NEW YORK TIMES ANALYSIS

1   Mr. McCarthy is credited with creating the term artificial intelligence in 1955, in a proposal for the Dartmouth conference.

2  The Times did not cover the Dartmouth workshop at all, let alone on the front page.

3  More recent online articles refer to the conference as starting “a new field of science.”

4  The Times could not find a reference to Mr. McCarthy ever saying this phrase, though it includes common ways to describe A.I.

Microsoft’s Bing cited its findings to a realistic-looking web address on The Times’s website:

MICROSOFT’S BING

Based on web search results, the earliest mention of artificial intelligence in the New York Times seems to be from a book review published on March 5, 1956. 1 The book is called “The Human Use of Human Beings” 2 by Norbert Wiener, and the reviewer is John R. Pierce. 3 The review mentions artificial intelligence as one of the topics covered by Wiener, who is described as “a pioneer in cybernetics”.:

https://www.nytimes.com/1956/03/05/archives/the-human-use-of-human-beings-cybernetics-and-society-by-norbert.html 4

NEW YORK TIMES ANALYSIS

1  The Times published a review of a book by Mr. Weiner one day earlier, on March 4, 1956.

2  This seminal book was published in 1950, not 1956. The March 4 review was of Mr. Weiner’s book “I Am A Mathematician: The Later Life of a Prodigy.”

3  The March 4 review was not written by John R. Pierce, a noted electrical engineer. It was written by John G. Kemeny, a mathematician.

4  The format of this URL is consistent with Times archives for articles in this period, but this exact webpage never existed.

According to The Times’s archives, all the chatbots were wrong. They cited articles that did not exist. And while coverage of early research on thinking machines dated to the 1930s, it wasn’t until 1963 that The Times first published an article with the phrase “artificial intelligence.”

“We released Bard as an experiment and want to be as transparent as possible about well documented limitations,” Jennifer Rodstrom, a spokeswoman for Google, said. “These are top of mind for us as we continue to fine tune Bard.”

Like Google, Microsoft and OpenAI say they are working to reduce hallucinations.

The new AI. systems are “built to be persuasive, not truthful,” an internal Microsoft document said. “This means that outputs can look very realistic but include statements that aren’t true.”

The chatbots are driven by a technology called a large language model, or L.L.M., which learns its skills by analyzing massive amounts of digital text culled from the internet.

By pinpointing patterns in that data, an L.L.M. learns to do one thing in particular: guess the next word in a sequence of words. It acts like a powerful version of an autocomplete tool. Given the sequence “The New York Times is a ____,” it might guess “newspaper.”

Because the internet is filled with untruthful information, the technology learns to repeat the same untruths. And sometimes the chatbots make things up. They produce new text, combining billions of patterns in unexpected ways. This means even if they learned solely from text that is accurate, they may still generate something that is not.

Because these systems learn from more data than humans could ever analyze, even A.I. experts cannot understand why they generate a particular sequence of text at a given moment. And if you ask the same question twice, they can generate different text.

That compounds the challenges of fact-checking and improving the results.

Bard said in one chat:

GOOGLE’S BARD

The earliest mention of artificial intelligence in the New York Times was in a February 19, 1950 article titled “Thinking Machines.”1 The article, by Walter Sullivan,2 reported on a meeting of the American Association for the Advancement of Science,3 where a number of scientists discussed the possibility of creating machines that could think. Sullivan wrote that “the possibility of creating machines that can think is no longer a remote one.”4 He quoted one scientist as saying that “we are on the threshold of a new era in which machines will be able to do many of the things that we now do ourselves.”

The response above has been truncated for length.

NEW YORK TIMES ANALYSIS

1  The Times published an article in November 1950 called “‘Revolution’ is Seen in ‘Thinking Machines.’”

2  Mr. Sullivan was a long-time science reporter at The Times.

3  The November 1950 article was about the meeting of a different group, the Society for the Advancement of Management.

4  This phrase does not appear in The Times archives from that era.

Then Bard said in another chat:

GOOGLE’S BARD

The earliest mention of artificial intelligence in the New York Times was in a 1955 article titled “Machine That Thinks”. The article described a computer program that could play checkers 1 at a human level. The program, developed by Arthur Samuel, was able to learn from its mistakes and improve its play over time. The article was written 2 by Herbert Goldstine, 3 a mathematician who worked on the Manhattan Project...The article was published on July 21, 1955, and it was met with mixed reactions

NEW YORK TIMES ANALYSIS

1  Mr. Samuel started programming his checkers player in 1952, and it debuted on television in 1956.

2  News articles in The Times are written by journalists, not subjects of the news.

3  Herbert Goldstine appears to be an inaccurate reference to Herman Goldstine, a noted computer scientist.

Companies like OpenAI, Google and Microsoft have developed ways to improve the accuracy. OpenAI, for instance, tries to refine the technology with feedback from human testers.

As people test ChatGPT, they rate the chatbot’s responses, separating useful and truthful answers from those that are not. Then, using a technique called reinforcement learning, the system spends weeks analyzing the ratings to better understand what it is fact versus fiction.

A newer version of ChatGPT called ChatGPT Plus, which is available for a $20 monthly subscription, consistently avoided answering the question about the first mention of artificial intelligence in The Times. This could be the result of reinforcement learning or other changes to the system applied by OpenAI.

CHATGPT PLUS

I am unable to browse the internet or access specific archives directly. However, I can provide you with some general historical context on the topic of artificial intelligence.

The New York Times, being a prominent newspaper, likely covered the topic soon after its inception, with articles in the late 1950s or early 1960s. Early mentions of artificial intelligence in the paper would have discussed the hopes and challenges of AI research, as well as the potential impact of AI on society.

Microsoft built its Bing chatbot on top of OpenAI’s underlying technology, called GPT-4, and has layered on other ways to improve accuracy. The company uses GPT-4 to compare the chatbot’s responses with the underlying data and rate how the model is performing. In other words, Microsoft uses the A.I. to make the A.I. better.

The company also tries to improve the chatbot’s responses with help from its traditional internet search engine. When you type a query into the Bing chatbot, Microsoft runs an internet search on the same subject and then folds the results into the query before sending it on to the bot. By editing the query, said Sarah Bird, a leader in Microsoft’s responsible A.I. efforts, the company can push the system to produce better results.

Google uses similar methods to improve the accuracy of its Bard chatbot. It uses human feedback to hone the system’s behavior, and it “grounds” the system using information from the company’s search engine, said Eli Collins, a vice president of research at Google.

Microsoft does not check the bot’s responses for accuracy in real time, Ms. Bird said, though it is researching how to do that. It checks the accuracy of a small portion of results after the fact and then uses that analysis.

But becoming more accurate may also have a downside, according to a recent research paper from OpenAI. If chatbots become more reliable, users may become too trusting.

“Counterintuitively, hallucinations can become more dangerous as models become more truthful, as users build trust in the model when it provides truthful information in areas where they have some familiarity,” the paper said.

Steve Lohr and Nico Grant contributed reporting. Jack Begg and Susan C. Beachy contributed research.

Source

Now, it's time to open up some discussion about the state of PC gaming in 2023. Many of us thought we would see the end of poor-quality games hitting the shelves. Yet following on from the review of The Last of Us: Part 1, which launched last month with an array of performance-impacting issues, I've come to the conclusion that even developers seem to not actually care about the quality of their PC ports, and only resort to Twitter statements about how they strive for better and will release patches in coming weeks/months to improve the situation.

But why is this? Why are many big releases being delayed from their original launch date (The Last of Us: Part 1) with the promise of a game in its "best possible shape", only to then release with lots of issues anyway?

It also feels like that a certain section of the online community is actively normalizing these technical problem-laden releases, by advocating gamers upgrade to the latest and greatest hardware to take advantage of brute forcing one's way through an optimization-impacted game. It does not help that a fair number of gaming outlets headline articles about memory requirements getting higher and higher, when the reality is that these are steep due to poor optimization at a game engine level.

Take for example the above screenshot of Cyberpunk 2077 where the engine is well optimized for a broad range of PC configurations. On my RTX 4090 with 12700KF CPU and 64GB of DDR4 RAM, I see under 5GB of system RAM usage, with VRAM usage under 12GB. Consider that this is using the 'Ultra Path Tracing' graphics preset with DLSS Quality and Frame Generation and SSR set to the highest Psycho option. The fact that a large open-world game like this with a bustling city with thousands of NPCs is capable of this sort of memory utilization whilst still having a healthy workload balanced across all CPU cores and making use of the GPU effectively proves that it is entirely possible to have a modern game engine run its best across the board.

That's not to say that Cyberpunk 2077 is a perfect benchmark for what modern games should be. It too launched with countless technical issues which took the good part of two years to resolve.

On the one hand, I can see where part of the blame lies with poor PC ports, and that's with us gamers. If we keep buying and pre-ordering new games without sampling a demo or being able to read reviews published from recognized outlets before launch, then many developers will keep launching shoddy ports that require months of patching regardless of what negative publicity says.

On the other hand, if we don't buy new PC titles due to the above, then developers could alienate PC as a platform for launch because the sales aren't there. It ends up being a catch-22 because this would not satisfy shareholders with a vested interest in sales and meeting release schedules.

I was in the lucky position of having had three generations of higher-end GPUs since Cyberpunk 2077 launched in 2020. An RTX 2070 Super, RTX 3080 Ti, and now, the RTX 4090. I have seen how games like Cyberpunk 2077 launch in a bit of a mess. Only those with brute force capability are able to power their way through to passable framerates, that is if the traversal and shader stuttering can be tolerated in most cases.

With Star Wars Jedi: Survivor, it is clear the game, story, and combat mechanics are very good, but the technical issues are simply too jarring and unacceptable. You would think that seeing the negative publicity that these big games get would ensure that games yet to be released undergo strict testing befitting of a polished and optimized title, but here we are.

The major issue surrounding Jedi: Survivor, is that the engine simply is not optimized for acceptable GPU and CPU utilization. On my i7 12700KF and 4090, the first three hours of the game sat between a nominal 60-75 fps at 3440x1440 with or without ray tracing enabled. Enabling FSR does absolutely nothing for framerates, yet in turn degrades the image quality on moving assets as shown by Alex at Digital Foundry.

Then, after I reached Koboh, the frame rate rose to 90-120 fps on average, with GPU utilization sitting comfortably at 90-99% - My first thoughts were that the performance issues may only have been in the first 3 hours of the game, but this was soon squashed. As soon as I reached the town at the bottom of Koboh, things took a downward turn with GPU utilization not exceeding 65%, and frame rate falling to the low 30s, on the world's most powerful gaming graphics card currently... Ouch. That's before I even mention the traversal stuttering during gameplay and frame time spikes that happen simply by moving the mouse around the settings menu.

What's even more puzzling is that in 2019, Jedi: Fallen Order was also released with performance issues. It seems that no lessons have been learned.

EA's statement about going through extensive testing to ensure that there are no adverse issues from patches is also laughable because it contradicts the very situation they are currently in with the game. If extensive testing is important, then why are we having this discussion with a game that is in an unfit state for launch in the first place?

It would be interesting to hear our readers' thoughts on this. Holding developers to account through online outcry for these issues seems to not be working, and time and time again we find ourselves sitting in the same seat talking about it. There have been no less than six big titles that have come out in the last 12 months with major performance issues at launch for the PC version, with developers then promising day 1 patches and beyond to iron things out,

in the case of Cyberpunk 2077, CDPR made things right, but it took a couple of years to get there. Other games like Uncharted still have bugs such as mouse camera panning judder, the same bug which plagued The Last of Us: Part 1 recently, and Naughty Dog managed to fix that within weeks after a heavy public slaughtering.

Uncharted, however, remains forgotten.

What I think would please many of us are game demos or early access through Steam or other storefront. Heck, if EA wants to get more people onto their EA Play/Origin platform, then early access where gamers can submit issues long before launch would be hugely popular, and it would make us feel like we were part of the process by the end. A demo (or early access) gives ample time to gauge what performance is like, and, if released a good amount of time before launch, gives enough time to apply pre-launch patching should bugs be reported.

A demo does mean extra work for developers, but they are doing extra work now anyway weeks/months/years after launching a game, so it's probably wiser to go the demo route surely.

Alternatively, they could also test games on the most popular hardware configurations in circulation. The Steam Hardware Survey is updated regularly and outlines what PCs the vast majority of gamers are playing. It's an invaluable resource for developers, yet it feels like no one other than gaming websites seems to make use of it.

Plus, not all review outlets can be relied upon either, so the demo/early access route does genuinely seem to me as the best step to take.

Thoughts?

These aren't the PC ports you're looking for - The current state of AAA PC gaming

The press release didn't mention the specific stock market trading board, but Reuters reports that it will be for the tech-heavy NASDAQ group. It added that Arm is expected to raise between $8 billion and $10 billion when it launches its IPO.

Right now, the current US stock market situation is highly volatile, thanks to what appears to be a looming recession. While some tech companies like Microsoft, Amazon, and others have reported solid financial results for the first quarter of 2023, it remains to be seen if that performance will continue through the rest of the year.

Arm and its parent company, Japan's SoftBank, are hoping that this IPO will fuel some new investment. In February 2022, Softbank confirmed that a proposed deal that would have seen Arm sold to NVIDIA for $40 billion was called off after it received resistance from US federal regulators.

Earlier this year, it was reported that Arm would start developing its own chip, rather than simply create designs that would later be used and made by other processor creators. The report claims that such a chip would be made to demonstrate Arm's designs, but it reportedly would not be made to be sold as a chip in devices sold to the public.

Arm has filed papers to register for a US stock market launch this year

It was the ultimate chip war that never was: German officials denied that Berlin planned to stop exporting specialty chemicals for chip fabrication, Reuters reported on April 27 – a day after Bloomberg News claimed that the government of Olaf Scholz “was in talks” on the subject, presumably under prodding from Washington.

The stock prices of BASF and Solvay, the largest makers of the specialty products, plunged on Thursday after the Bloomberg report appeared but recovered sharply on Friday after the government’s denial.

More than a dozen chemicals including acids, bases and solvents are indispensable to etching microcircuits onto silicon wafers, and an interruption of supplies would cripple China’s fabrication capacity. Restricting these chemicals would escalate the chip war far beyond the scope of the Biden Administration’s October 7 controls on semiconductor equipment and design software for the most advanced chips, used in high-end smartphones, servers and artificial intelligence applications. The target would include all chips including so-called mature processes.

In related news, US Commerce Secretary Gina Raimondo announced on April 26 that the US will consider banning exports to Chinese cloud-computing companies, including Alibaba as well as Huawei. Raimondo responded to a letter from a group of Senators led by Bill Hagerty (R-TN) warning that Chinese cloud computing threatens US “national security and economic security.”

The Biden Administration is also expected to ban US investments in yet-to-be-announced high-tech industries in China.

Republicans on the House Foreign Affairs Committee meanwhile are wrapping up a three-month investigation of the Commerce Department’s enforcement of chip controls and have offered legislation that would give the Defense Department responsibility for enforcing the ban. Gregory Allen of the Center for Strategic and International Studies argued in congressional testimony April 13 that “China’s export control evasion activities are significant and growing. My primary recommendation is that Congress focus on concrete strategies to tighten this enforcement and shore up remaining gaps that risk allowing China to close the AI gap.”

Washington is scrambling to restore credibility to its effort to contain China after a streak of Chinese diplomatic victories, including the Beijing-mediated restoration of diplomatic relations between Iran and Saudi Arabia, French President Emmanuel Macron’s visit to Beijing in a show of independence from Washington and the visits of Brazilian President Lula and Malaysian President Anwar Ibrahim.

Ukraine President Zelensky’s telephone call with Xi Jinping and the dispatch of a special Chinese envoy to Ukraine, moreover, raise the prospect that China will pick up the pieces in Ukraine, after a drumbeat of damaging Pentagon leaks revealed how badly America stumbled.

In the advent of the 2024 presidential election, the Democratic administration is sensitive to Republican claims that it is too easygoing towards China.

China’s economic success in the Global South threatens to lure key countries out of the American orbit. As of March, China’s exports to ASEAN countries rose 35% year-on-year and exports to Central Asia (including Turkey and Iran) rose 55%, Asia Times reported April 25. China now exports more to the Global South than it does to developed markets.

Washington’s controls on the export of high-end chips and chip-making technology to China, announced October 7, 2022, were intended to deny China access to cutting-edge hardware that supported the most advanced AI applications. By contrast, the Commerce Department has shown flexibility in allowing semiconductor equipment manufacturers to ship machines that produce mature chips.

The Biden Administration adapted its strategy from a September 2022 report by the Special Competitive Studies Project, chaired by former Google CEO Eric Schmidt. This offered a response to China’s growing military power as imagined by Silicon Valley software venture capitalists: An “Offset-X strategy” including “distributed and networked operations, human-machine collaboration, human-machine teaming, primacy in software-centric warfare, resilience and greater technological interoperability and interchangeability and partners.”

Eric Schmidt speaks during a National Security Commission on Artificial Intelligence (NSCAI) conference November 5, 2019, in Washington, DC. Photo: Asia Times files/ Alex Wong / Getty Images via AFP

That was a Silicon Valley futurist’s view of warfare, unrelated to military technology that will prevail for the foreseeable future. Both the US and Chinese military use older-generation chips for sensing, targeting and processing information. The older chips are more robust and easier to harden, as a February 2022 Rand Corporation study explained.

The Biden Administration gravely underestimated the power and importance of mature chip technologies (14 nanometers and higher), which comprise 95% of the global chips market and power 5G infrastructure, industrial productivity applications, and other so-called Fourth Industrial Revolution technologies. Semiconductor fabricators depend on mature chips for most of their revenues, and China’s massive investment in a domestic supply chain threatens to erode the financial base of the whole Western chip industry, as Dimitri Alperovitch of Silverado Incubator has observed.

One problem is that cutting off the Chinese market may have devastating consequences for the revenues of Western high-tech companies. The Atlantic Council warned in a March 2023 report, “While the steps taken by the Biden administration to constrain China’s progress in producing cutting-edge semiconductors appear calibrated to avoid widespread industry disruption, the policy has painful consequences that cannot be downplayed….the bottom-line impact may be felt in terms of what the industry terms a ‘significant loss of scale’ that could yield fewer resources for R&D and new investments…. It is essential that the semiconductor industry – and US allies, as discussed below – have a voice in assessing the potential impact of additional proposed constraints. Communication is essential to avoid unintended consequences.”

To make matters worse, American sanctions on the sale of high-end chips to China are extremely difficult to enforce. To comply with sanctions Nvidia reduced the clock speed on a variant of its GPUs, the standard for high-end servers, while selling substantially the same product to China.

In addition, the global market in chips is so complex and opaque that Chinese companies can buy whatever they want on the gray market, and the enforcement capacity of the Commerce Department is woefully inadequate to prevent this, Allen stated. Chinese sources say that high-end GPUs are freely available at a 10% premium to the going price.

Chinese commentators compare the chip war to China’s civil war. By concentrating on low-end chips, and undercutting the prices of Western manufacturers, “Observer” columnist Chen Feng says, China will “encircle the cities from the countryside,” a reference to Mao’s successful military strategy during the Civil War of the 1940s. The United States “can only go to Menglainggu [the site of a decisive 1947 Communist victory) by relying on high-end chips.” As noted, analysts like Alperovitch have already flagged the danger.

On March 27, Huawei announced that it had developed its own chip design software for 14 nanometer and wider mature nodes; if true, that would represent a major step towards Chinese independence from US design firms Cadence and Synopsis, which have had a near monopoly on the technology.

As the Atlantic Council suggested, the Biden Administration appears to have given US firms considerable leeway in exports to China. LAM Research, a leading US manufacturer of semiconductor equipment, predicted an increase in sales to China during the remainder of 2023, after receiving “clarification” of export rules from the Biden Administration. The Dutch chip lithography giant ASML also projected an increase in Chinese sales this year, and its CEO Peter Wennink stated that “the mature semiconductor space is very important and needs to grow. And this is where China is very strong.”

The reported Berlin ban on semiconductor chemicals might have misfired, but it was the first salvo of what will be a long tech war of attrition.

Source

Uninstall Microsoft Store DLC in Windows 11

1. Launch the Settings app by pressing Win + I or another method you prefer. You can also right-click the Start menu and select "Settings." Alternatively, press Win + R and type "ms-settings:appsfeatures-app" without quotation marks. The command will get you directly to the list of installed apps on your system.
2. In the Settings app, go to Apps > Installed apps.

    

   
3. Find the app which DLC you want to uninstall. Another option is to click the search bar and type the DLC's name.
4. Click the three-dots button next to your game. Select "Advanced options."

    

   
5. Scroll down on the next screen and find the "App add-ons & downloadable content" section. Here you can uninstall game DLCs downloaded from the Microsoft Store. If the game has a lot of additional content, use the search bar or sort the list with filters, such as drive, name, size, and install date.
6. Click the DLC you want to uninstall and then press the Uninstall button.

    

   
7. Wait for Windows 11 to remove the selected content. Note that the process will take some time, depending on your computer's speed or DLC's size.

Note that some apps have built-in stores and repositories for managing downloadable content. Usually, those do not appear in the Settings app, like DLCs from Minecraft's Marketplace. Therefore, to uninstall such a DLC, check the corresponding section inside the specific game.

How to uninstall game DLCs downloaded from the Microsoft Store in Windows 11?

Ironically, this announcement came nearly 40 years to the day that the first Microsoft Mouse was sold. That happened on May 2, 1983, and it would be the first of many, many PC mice of all different designs from Microsoft.

The Microsoft Hardware division was formally formed one year earlier in 1982, sometime after the company started selling the Z80 SoftCard add-in card for the Apple II in 1980. However, by 1982, other companies like Mouse Systems, IBM, and Logitech had launched mice products for IBM-based personal computers. Microsoft decided it wanted to enter that small but growing market as well.

When the mouse was first released on May 2, 1983, it sold for a whopping $195. That was a massive expense for a PC accessory back then, and today that price would be more than enough to get a very fancy wireless gaming mouse from Logitech or Razer.

With its distinctive pair of green buttons, it's no wonder that the first Microsoft Mouse got the nickname "The green-eyed monster". There was no "plug and play" USB connection here. This first mouse product from Microsoft had an InPort ISA interface that also required that the PC have a Microsoft bus card inside. It used a steel ball for its sensor and had a slightly curved design that made it fit more comfortably in the hand compared to other mice products that looked and felt like the owner was moving a brick around. Of course, you could also see the cool Microsoft logo on top of the mouse.

While the mouse was first sold in May 1983, the accessory was later bundled with the first version of Microsoft Word, which launched just a few months later in September 1983 (and you can bet we will be writing about that 40th anniversary when the time comes). ToastyTech has some more info on the software that came with the mouse. It included "PIANO.EXE" a virtual piano program. along with "DOODLE.EXE" a drawing program.

Other applications that were included in the first Microsoft Mouse were the DOS version of Notepad. There was also "LIFE.EXE." No, this was not the classic board game, but a game that tried to simulate the rise and growth of microorganisms.

If you are curious, you can download the PDF of the original printed manual for the first Microsoft Mouse on the company's website. It's actually a fascinating read and a true look back at PC technology from 40 years ago. You can even download the mouse's software with the bundled Microsoft Word for DOS.

Even with including the Word software with the mouse, Microsoft's first PC accessory product didn't sell all that well. This Day in Tech History says that Microsoft actually made 100,000 units of the original mouse, but only sold 5,000 of them.

It would take the second version of the Microsoft Mouse released in 1985 before the company's mouse products started to take off. That second mouse evolved the original's curved design but ditched the distinctive green buttons for grey ones. The company would continue to evolve its PC mice products over the decades, including products like the Intellimouse, the Arc Mouse, and more.

It now seems like Microsoft's PC accessory hardware business is at a crossroads with this week's announcement that it would no longer offer any new devices under the Microsoft brand. The Surface division does continue to sell a version of the Arc Mouse. However, it remains to be seen if that part of Microsoft is interested in making any new mice devices that are made to work with all PCs.

Of course, there's also a ton of competition from other companies who are more than willing to innovate and evolve the PC mouse. However, it would be a shame if Microsoft, one of the true pioneers in this market, were to quietly go away.

When you purchase through links on our site, we may earn an affiliate commission.

A look back at the first Microsoft Mouse released nearly 40 years ago

Not so long ago, Intel had announced that the 14th-gen version of Intel Core processors, named Meteor Lake, will bring something new to the tech, something Intel has never done before.

Intel Meteor Lake CPUs will use a tile-based design. Where the processor will not be a single, monolith one, but instead will have smaller tiles inside it.

Simply put, it will basically have 4 tiles (5 if you count the base):

• The CPU tile, which will have P and E cores, among other things.
• The GFX tile, which will have the GPU, which is the graphics processor.
• The SoC tile, which will look after the RAM, display ports, media playback and others.
• The I/O tile, which will look after PCIe, SATA, USB, Ethernet and other connections.

The above tiles will be combined to make a single Meteor Lake CPU.

In this tile based architecture, Intel was forced to follow AMD, as this is something that AMD does with it’s Ryzen processors. However, AMD does it slightly differently, where it splits the two 8-core (max) CPUs into a chiplet form and also has an additional third SoC chiplet. Here, Intel is putting all the CPU cores inside a single tile instead.

But why do it in the first place. It’s because making smaller chiplets / tiles by splitting things is cheaper, for a start. Another thing is, there are rumors that just like how TSMC is making Intel ARC graphics card chips, they would make Intel GPU chips in Meteor Lake too.

But what about the Meteor Lake Release Date. Looks like Intel has given more information on it.

Intel Confirms Meteor Lake Coming In 2H23

In its quarterly earnings report, Intel has confirmed that Intel Meteor Lake CPUs are coming in the second half of this year.

It mentions it is based on Intel 4 process, which actually can be called 7nm, but it is similar to TSMC’s N4 process.

Additionally, it mentions that Intel is ramping up it’s production in this current second quarter of the year, so that it can release the chips by the second half of this year.

Why is all this a surprise. This is because there were rumors that Meteor Lake will not release this year, but instead the current-gen Raptor Lake is going to get a refresh. Now, Raptor Lake Refresh could happen, but Meteor Lake coming this year is a good thing.

Not only that, it mentions that the 15th-gen Intel Core CPUs, named Arrow Lake, are going to come next year, in 2024.

Not to forget, Intel claims Intel 3, Intel 20A (2nm) and Intel 18A (1.8nm) nodes are on track. Those unaware, starting 2nm, Intel will not use nm (nanometer) but a (angstrom) instead.

In all, Intel claims that it remains on track to achieve 5 process nodes in 4 years.

But in all this. There’s one thing. Intel Meteor Lake’s new tile architecture might not be broadly available on desktops.

Meteor Lake Desktop Might Have Only i3 And i5 CPUs

A couple of months ago, we shared how rumors are going around that Intel may or may not release Meteor Lake on desktop. It mentioned that if at all, it will be limited to 6 cores. Looks like there’s an update to it.

In a tweet, @SquashBionic has shared a photo which basically confirms the above.

The table in the photo confirms some things. First is, Meteor Lake is coming on desktop. Second is, it’s going to use a new socket, LGA1851. However, the biggest and most important is, it will have CPUs only up-to i5 and will be limited by 35W and 65W CPUs.

What does the above mean. It means that Intel Meteor Lake CPUs will have only i3 and i5 processors. With higher powered CPUs, like 125W using i7 and i9, being reserved for 15th-gen Arrow Lake CPUs.

This revelation, if true, basically confirms the previous rumors, which mention that the P-Cores in Meteor Lake will be limited to 6, which is only found in Intel Core i5 processors.

This is not the first time Intel has done something like this, though. For example, all the CPUs in the current-gen Intel Raptor Lake line-up are not actually based on Raptor Lake. Some CPUs in the i5 and i3 line-up are said to be based on previous-gen Alder Lake CPUs, just that they have an increased clock rate to appear faster.

Intel Meteor Lake Could Come With L4 Cache

The rumors about the Meteor Lake don’t seem to conclude, which is a good thing, as it seems to come with a lot of new features. One of them is L4 Cache.

In the patches released for Linux, Intel has basically confirmed that Intel Meteor Lake is going to have L4 cache.

But if that’s not an indication enough, it seems that some years ago Intel had filed a patent for the same. It mentions the use of L4 Cache in detail.

While L4 cache is not entirely new, Intel has used it previously for GPU inside the CPU and it’s mostly for performance reasons. The patent mentions a lot of things Intel is wanting to do with it.

But mainly, Intel seems to concentrating on the security improvements it can offer. It mentions that due to faster access time (compared to RAM), it can be used for pre-boot security purpose among other things.

This could mean that this extra cache is possibly not a counter to the AMD Ryzen X3D processors.

What’s interesting is that it seems that Intel will put this L4 cache on the base of the CPU, not inside any of the tile, at least from the looks of it.

Either way, all this is to be seen. Will Intel, which is known for delays, release Meteor Lake on time. Whether it comes on desktops. Will the new L4 cache, offer any performance benefits. Only time will tell all that.

Intel Confirms 14th-Gen Meteor Lake CPU Are Coming This Year

Windows security

Windows security was a hot topic in our news coverage this week. Microsoft released various PowerShell scripts to help IT admins identify if recent patches have mitigated security flaws in their Windows 10 and 11 environments. It also shared a timeline covering the next few months of Windows hardening for various protocols including Netlogon and Kerberos.

However, the firm continues struggling with existing security issues in its operating systems. It confirmed yesterday that the latest Patch Tuesday update has broken Local Account log-in on both Windows 10 and Windows 11, a workaround has been issued. It has also cautioned about interop issues in its Local Administrator Password Solution (LAPS) but has recently fixed most, if not all of them, too. IT admins can also leverage an unofficial third-party GUI tool to manage both versions of LAPS, at their own risk. Another problem related to Local Security Authority (LSA) protection in Windows Defender has been patched too.

Additionally, Microsoft revealed more details about the security features present in its revamped Teams desktop app. It might interest some of you to know that the firm is considering unbundling Teams from its Office suite of apps to appease EU regulators and also empower Teams users on Android to disable automatic firmware updates.

And if you like to mess around with Windows installations, you might want to check out AtlasOS, which is a third-party, lightweight version of Windows 10 that's supposedly better for gaming. It received updates during this week and is set to get back Windows Defender soon too, with the developer making a seemingly tall claim that AtlasOS will be even more secure than the official Windows 10. But if you'd rather mess with your existing installation at a smaller scale, you can check out Wintoys for Windows 11 instead.

Edge leakage

Edge and Microsoft's other browser-related efforts found themselves in the news a lot throughout this week, for both good reasons and bad. For starters, someone made a rather startling discovery that Edge is apparently leaking all the websites you visit in the browser to the Bing API, even if you don't use Bing. Microsoft is still investigating this claim and has promised swift action once its findings are complete. Brave has also decided to stop using the Bing API to serve search results and its search index is now 100% independent.

We also learned that Microsoft has started to insert ads for the new AI-powered Bing in the search bar if you visit Google Bard through Microsoft Edge. Bing Chat itself received a few updates this week to further decrease the disengagement rate and improve answers for traveling and cooking questions.

Talking more about browsers, Windows 11's Mica effect is slated to arrive in Chrome soon. Moreover, the latest Edge Dev 114 update has added several fixes and a shortcut for toggling Sidebar. A recent Canary build is also making it easier to tweak Windows 11-style rounded corners in Edge. In fact, Microsoft has announced that Edge users will soon be able to disable some unnecessary features as well.

There were other items to cover in the Microsoft 365 sector too. All Microsoft 365 apps and services will begin using the more neutral "cloud.microsoft" domain soon. Microsoft has also been improving its Microsoft 365 apps and services themselves in recent days. It has detailed all the improvements it made to Excel in April, added the Accessibility ribbon to Outlook for Windows, integrated the Today tab in the latest Skype Insider build, and is implementing a way to make it easier for Outlook mobile users to reply to emails on time.

Meanwhile, Microsoft Loop can now be tested via personal accounts on iOS and Android while Microsoft Designer is now available in full via a public preview. Something that will displease Windows Weather app customers though is that Microsoft has decided to infest it with MSN news content, which is ludicrous, to say the least.

The end of Windows 10

Microsoft announced earlier this week that version 22H2 is the last version of Windows 10. What this means is that while customers using the OS will continue receiving minor improvements and security updates until October 14, 2025 (and beyond if you're an LTSC customer), there will be no more feature updates. In addition, Microsoft has also announced that the LTSC version of Windows 11 will arrive in the latter half of next year. Windows 10 recently received April's non-security preview update with lots of improvements too.

Another thing being discontinued is Microsoft-branded mice, keyboards, webcams. The remaining accessories will only be sold until current supplies last, after which Microsoft will focus only on Surface-branded accessories.

In more end-of-support news, the Steam client has started "End of Life" alerts for some users as it will be dropping support for Windows 7, 8, and 8.1 on January 1, 2024. Roughly 2% of Valve's user base utilizes these three operating systems, which means that the vast majority won't be affected.

Windows 11 continued to be Microsoft's focus of attention when it came to operating system releases this week. Apart from the expected April non-security preview updates for Windows 11 version 21H2 and 22H2, there were a few Insider releases too. Windows 11 Beta build (KB5025303) fixed LAPS and ReFS issues, and added new Widgets. Meanwhile, Canary build 25352 brought a new widget picker experience.

Some may also find it interesting that Microsoft is looking to make Rust a part of the Windows 11 kernel due to the numerous advantages it offers in writing safer and more optimized code. But in terms of what's available generally right now, we have Phone Link for iOS and guidance for how to get Windows 11 updates faster. And if you're still on the fence about giving Windows 11, make sure to try out the Redmond tech firm's latest evaluation virtual machines for the OS.

Git gud

Microsoft faced probably its biggest roadblock yet in its bid to purchase Activision Blizzard when the UK Competition and Markets Authority (CMA) decided to block its deal, citing potentially reduced competition in the cloud gaming space. As Activision stock tanks, Microsoft has decided to appeal the decision, with president Brad Smith calling this initial outcome "bad for Britain".

Although another report claimed that Microsoft may close the deal without the U.S. Federal Trade Commission's (FTC) approval, that rumor was based on the UK CMA giving its blessing in the first place. Now that this hasn't happened, it's unlikely that the purchase will close anytime soon. For now, Microsoft's acquisition has been approved in Ukraine and the firm has signed yet another 10-year cloud gaming deal with a European firm too (Nvidia GeForce Now will continue to get Xbox games despite this setback too). Activision's CEO Bobby Kotick fully expects the CMA's decision to be reversed following the appeal.

In other gaming news, there have been a few Xbox Insider builds to fix game art issues and other bugs. And if you're eagerly awaiting Redfall, you can build themed Xbox controllers based on the game through Xbox Design Lab. Similarly, if you reside in London or NYC, you might want to check out the dedicated gaming zones that Microsoft has built with Samsung.

Coming over to game updates, Flight Simulator World Update XIII is now available, it targets Oceania and Antarctica. Quantum Break fans will also be pleased to know that the title is back on digital storefronts and is available on PC Game Pass for the first time too. Meanwhile, Grounded players can tackle wasps as a new foe, with the game also being Steam Deck verified now. And in a major win for accessibility, Turn 10 Studios has announced that it is working on tons of accessibility options for the next Forza Motorsport, allowing blind players to enjoy the game as well.

Finally, talking about the deals and promotions currently going on, Xbox Free Play Days is offering four games, namely Crusader Kings III, Hell Let Loose, Don’t Starve Together, and Leap. May's Games with Gold have also been announced and they are Hoa and Star Wars Episode I Racer. And if you're on the lookout for games on the cheap, check out the latest Xbox Deals with Gold and this Weekend's PC Game Deals, curated by our News Editor Pulasthi Ariyasinghe.

Dev Channel

• Microsoft's Q3 2023 results show growth but Windows, Xbox, and Devices numbers continue to decline
• Rufus 4 is out with improvements and a default 64-bit executable
• Rufus alternative WinToUSB has received "Windows 11 Lite" support
• Latest WingetUI has a new UI, faster loading performance, and lots more
• Microsoft has decided to continue working with private Russian companies not under sanctions
• Fluent Emoji Gallery app is now available with access to Microsoft's three emoji styles
• Apple Music Preview for Windows has finally netted media keys and lyrics support
• Microsoft-backed BSA has called for federal rules and guidelines on AI
• You can now create Power BI reports directly in Jupyter Notebooks

Under the spotlight

News Reporter John Callaham took a look back at Windows Server 2003, which launched 20 years ago this week. It was the server version of Windows XP and is reportedly still being used in some enterprise environments. Read more about its interesting, but brief, history here.

John took another trip down memory lane this week as he reminisced about Microsoft's first PC game, simply called Microsoft Adventure.

Meanwhile, News Reporter Taras Buria penned a guide about how you can use a Windows laptop or a tablet as a secondary monitor.

Finally, forum member Adam Bottjen published his latest Tech Tip Tuesday guide explaining how you can quickly fill an online form with ease.

Logging off

Our most interesting news item for this week involves someone building a ChatGPT client in Java that is capable of running on Windows 98, ME, 2000, XP, Vista, 7, 8, 10, and 11. It's called JavaGPT and weighs in at just 6MB. While it's doubtful that it has any real-world use considering that you can just access the service from a browser, it's interesting to see experiments like these from time to time. Find out more details about the open-source Java utility here.

Microsoft Weekly: Windows security, Edge leakage, and the end of Windows 10

Publisher Paradox Interactive recently unveiled the final expansion for Cities: Skylines, called Hotels and Retreats. Developer Colossal Order announced this expansion would be the last for Cities: Skylines, after eight years of post-launch content, as it continues to work on a sequel set for release later this year.

Hotels and Retreats will allow players to build various hotels and tourist attractions in their cities, boosting their economy. This comes after eight years of post-launch content, as developer Colossal Order gears up for the sequel, which is set to release later this year.

Check out Cities: Skylines' last expansion trailer below.

What to expect from Hotels and Retreats?

The expansion will feature a range of hotel buildings, from hostels to luxurious resorts, which players can add to their cities on different budgets. Players will start with one-star hotels and unlock new options as their business grows.

To support the hotels, new structures such as parks, restaurants, playgrounds, and cafés will also be available. Matching hotels with good locations will be crucial, as different locations will attract different tourist groups. The expansion will feature five new maps, including three European-inspired maps, one tropical map, and one temperate map, giving players a range of options for their hotel locations.

Additional content creator packs

In addition to Hotels and Retreats, three content creator packs are also coming out, which include growable industrial buildings from various eras, Japanese railroad scenery, and growable buildings and props inspired by Brooklyn and Queens.

Two new radio stations

Furthermore, two new radio stations will be available, featuring two hours of piano music and '90s pop.

Hotels and Retreats is the name of the Cities: Skylines' last expansion

The legacy of Cities: Skylines

The game's legacy is truly impressive, as it filled the void left by the failure of SimCity when it came out in 2015. In the years since its release, it has become the go-to city builder game for many players.

However, the competition is fierce now, with many similar games on the market, but the reputation of the original game should give the sequel a solid foundation. We will see how well the Cities series' new game will compete with others.

Cities: Skylines will bid us farewell on May 23rd with Hotels and Retreats

The new Xbox Insider Beta build is 2305.230426-2200. Here is the changelog:

My Games & Apps

• Fixes to address an issue where installing or updating titles could remain in the Queue ‘finishing things up’ longer than expected.
• Fixed an issue where Remote Tools for Microsoft Edge would repeatedly prompt for an update.

System

• Various stability and performance fixes.

The Xbox Insider Alpha Skip-Ahead build is 2308.230425-2200. Here is the changelog:

Fixes Implemented

Thanks to the hard work of Xbox engineers, we are happy to announce the following fixes have been implemented for this build:

System

• Various updates to properly reflect local languages across the console.
  • Note: Users participating in Preview may see “odd” text across the console, for more information go here.

Microsoft released new Xbox Insider builds earlier in the week for the Omega, Delta, and Beta rings. All of them address an issue where some steering wheels connected to an Xbox console unexpectedly lose force feedback.

Microsoft releases new Xbox Insider Beta and Alpha Skip-Ahead builds with bug fixes

Moreover, the company seems to be very optimistic about its future CPU plans, according to its press release:

Intel continues to be on track to meet its goal of achieving five nodes in four years, with two of the five nodes nearly complete. Intel 7 is in high-volume manufacturing and CCG's Meteor Lake product on Intel 4 is ramping production wafer starts for an expected launch in the second half of 2023. Intel 3, Intel 20A, and Intel 18A remain on track.

While on the surface, hearing that the Meteor Lake CPU design is still coming out this year is good news, PC Gamer notes that it might be released first as a low-power laptop product, rather than a high-end desktop or laptop CPU. In fact, Meteor Lake might not be released as a desktop chip at all. Current rumors claim that the next big 14th Gen Intel Core desktop chip might actually be a minor refresh of the current Raptor Lake-S design.

2024 is when we might see a big desktop CPU architecture launch, with Arrow Lake, which is supposed to use the previously mentioned Intel 20A node. That will be a true next-gen desktop release from the company, if it keeps its current schedule. If you are wondering about performance, Intel could be bringing a new L4 "Adamantine" or "ADM" cache, which could boost booting speeds.

Intel still says Meteor Lake CPUs are coming out in late 2023, but there might be a catch

Nuke-launching AI would be illegal under proposed US law

Intel has released its financial results for the first quarter of 2023, and the numbers aren't good: The company brought in $11.7 billion, a drop of 36 percent from Q1 of 2022. Because of the decline in revenue, Intel lost $2.8 billion, the biggest loss in the company's history. Intel's two biggest divisions—its client computing group that sells products made for end users and its data center group that sells server products—were down 38 and 39 percent, respectively.

If there's any kind of silver lining for Intel in the earnings report, it's that the company did slightly better than it expected to do this quarter; three months ago, Intel predicted revenues of 10.5 to 11.5 billion for Q1.

The company has instituted layoffs and has cut pay for managers and executives to help stem losses. It has also canceled some product lineups recently, like its Blockscale series of Bitcoin-mining ASICs.

Some of Intel's problems are affecting nearly every company that makes PCs or PC components; there has been a large overall decline in the consumer PC business as people continue to use equipment they bought during the pandemic. Analysts at IDC say that PC shipments are down nearly 30 percent from the same time last year. Samsung just posted a 95 percent decline in profits because of an oversupply of memory and storage chips. Companies like Nvidia, Micron, and even Apple have all posted declines in revenue in recent quarters.

But some of Intel's problems are unique. A big one is the company's years-delayed range of next-generation "Sapphire Rapids" server CPUs, which has given AMD its best opportunity in years to cut into the lucrative server CPU market with its EPYC chips. It's one reason why AMD's recent financial results have been a bit rosier—revenues from its data center division are helping to make up for lower consumer demand for its CPUs and GPUs.

Intel's Arc GPUs also arrived months later than planned and could only compete with Nvidia and AMD's less-profitable entry-level and midrange graphics cards (they also had lots of driver problems early on, though those have gradually improved). Intel originally had a separate graphics division (AXG) that reported its revenue separately from the other groups within Intel, but since AXG was dissolved and folded into the client computing and data center groups, we don't know how much money its GPU efforts have made (or lost).

Under CEO Pat Gelsinger, Intel is trying to turn things around by fixing these execution problems while also letting other companies make their chips using Intel's foundries. This so-called "IDM 2.0" strategy was announced over two years ago, but the long lead times for chip designing and manufacturing mean we're still waiting to see whether it will pay off. The company's foundry services revenue was down 24 percent year over year, but at $118 million, it's also just a tiny fraction of Intel's revenue; Gelsinger reiterated in Intel's press release that he sees it as a "$1 trillion market opportunity" in the long run.

"We are prioritizing the investments needed to advance our strategy and establish an internal foundry model, one of the most consequential steps we are taking to deliver on IDM 2.0,” wrote Intel CFO David Zinsner.

The company expects its pain to continue next quarter. Intel projected revenues of between $11.5 and 12.5 billion for the Q2 of 2023, which would be another big drop from the $15.3 billion the company reported in Q2 of 2022 (which itself was already a 22 percent drop from Q2 of 2021).

Source